China to tighten ethical review of AI patents

China will strengthen ethical examination for artificial intelligence in the patent review process to ensure AI technology complies with laws, social morality and public interests, the country's top intellectual property regulator said on Friday.

In a press conference held by the China National Intellectual Property Administration (CNIPA), senior CNIPA official Jiang Tong said the newly revised patent review guidelines, which will take effect on January 1, 2026, establish a dedicated section on AI and big data for the first time.

The section states that the implementation of technical solutions related to AI, such as data collection and rule setting, must comply with legal requirements, social ethics, and public interests, Jiang added.

It also specifies requirements for writing descriptions in scenarios such as model construction and model training, and refines the criteria for determining adequate disclosure, addressing potential issues of insufficient disclosure of technical solutions that may arise from the black-box nature of AI models.

At the press conference, another senior CNIPA official, Liang Xinxin, said the IP rights development plan for the 15th Five-Year Plan period (2026-2030) will continuously deepen international cooperation and exchanges.

It will also convey China's commitment to protecting the IP rights of both domestic and foreign-invested enterprises equally, foster a favorable business environment for all types of enterprises, and promote dual circulation of domestic and international markets, he added.